CR网络中基于能量检测的自适应门限改进算法研究

发布时间：2018-05-01 23:45

本文选题：认知无线电 + 频谱检测　；参考：《宁夏大学》2016年硕士论文

【摘要】：最近几年,信息化技术发展迅猛,人们对频带资源的需求度不断攀升,有限的频谱数量已不能满足亿万用户及设备的需求。另外,提升匮乏频带的利用率也成了人们当前亟待解决的难题。认知无线电频谱感知技术由于具有迅速、准确检测空闲频谱的能力,成为提高频谱利用率的一种新兴手段。本文主要对频谱感知的判决门限进行了研究和改进。首先对几种传统单用户频谱感知算法进行阐述、对比,同时,介绍了能够缓解多径、阴影以及终端隐蔽等现象的多用户协作频谱感知,该技术在弥补单用户感知缺陷的同时也加大了网络开销。由于传统算法大多根据虚警概率设置判决门限,使得门限无法适应快速变化的通信环境。基于此,本文研究并改进了两种判决门限自适应调整的频谱感知优化算法,并将其运用于协作频谱感知中,以减小协作用户数,降低网络开销,提高检测性能。各认知用户根据感知信噪比的差异动态调整门限阈值,更加实时有效地提升检测精度。前一种算法通过计算单用户和整个认知网络的最小错误概率,得出对应的门限值,并将检测中授权用户占用频段和不占用频段的概率及时间随机化,使优化算法更贴近于实际。后一种增强型自适应门限感知算法利用检测参数β对检测概率和虚警概率所对应的门限值进行动态调整,并对来自各方的授权信号进行多次检测,选择衰落最小的信号进行判决,令判决结果更加可靠。
[Abstract]:In recent years, with the rapid development of information technology, the demand for frequency band resources is increasing, and the limited number of spectrum can no longer meet the needs of hundreds of millions of users and equipment. In addition, improving the efficiency of the scarce frequency band has become a difficult problem to be solved. Cognitive radio spectrum sensing technology has become a new method to improve spectrum efficiency because of its rapid and accurate ability to detect idle spectrum. In this paper, the decision threshold of spectrum sensing is studied and improved. Firstly, several traditional single-user spectrum sensing algorithms are described and compared. At the same time, multi-user cooperative spectrum sensing, which can alleviate the phenomena of multipath, shadow and terminal concealment, is introduced. This technique not only makes up for the defect of single user perception, but also increases the network overhead. Because most of the traditional algorithms set the decision threshold according to false alarm probability, the threshold can not adapt to the rapidly changing communication environment. In order to reduce the number of cooperative users, reduce the network overhead and improve the detection performance, two kinds of spectrum sensing optimization algorithms with adaptive adjustment of decision threshold are studied and improved in this paper. Each cognitive user dynamically adjusts the threshold threshold according to the difference of perceived SNR to improve the detection accuracy in real time and effectively. By calculating the minimum error probability of the single user and the whole cognitive network, the former algorithm obtains the corresponding threshold value, and randomizes the probability and time of the authorized user occupying and not occupying the frequency band in the detection, so that the optimization algorithm is closer to the reality. The latter adaptive threshold sensing algorithm dynamically adjusts the threshold values corresponding to the detection probability and false alarm probability by using the detection parameter 尾, and detects the authorization signals from all sides several times. Choosing the least fading signal to judge makes the result more reliable.
【学位授予单位】：宁夏大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TN925

【参考文献】